Research On Fault Diagnosis And Health Management Of High-Aspect-Ratio Unmanned Aerial Vehicle Wings

For the high-aspect-ratio fixed-wing UAV wings,the wing will inevitably fail due to the complex operating conditions such as heavy load,large disturbance,and strong impact.The maintenance cost of the wing is increasing year by year due to the lack of an effective fault prediction system.However,the fault prediction of the wing is often characterized by complexity,weakness,coupling and complex causality,which makes the traditional fault prediction method not achieve good results.As a newcomer in the field of intelligent fault diagnosis,deep learning can independently explore representative diagnostic information hidden in the original data and directly establish an accurate mapping relationship between the original data and the running state.Therefore,the application of deep learning methods to the fault prediction of the wing has broad application prospects and high commercial value.In this paper,a series of health management studies on strain fault diagnosis,fault prediction and full life cycle prediction for composite aspect ratio UAV wings are performed,mainly including the following research contents:1.Collecting wing strain data: Experimenting with real experimental scenes,adopting the method of installing sensors on the inner and outer skins of the composite wing,directly collecting the strain data generated by the wing during the loading process to ensure the data is authentic and reliable.2.Single point fault diagnosis: Analyze and learn the collected wing strain data through deep confidence network,find and establish the accurate mapping relationship between wing strain data and wing state,and achieve accurate fault location and diagnosis.3.Multi-point coordinated fault prediction: collaborative analysis and learning of multiple sensor data,establishing a coordinated trend of multiple sensors corresponding to the fault occurrence so that fault location is more accurate.4.Multi-point coordinated health management: Through the analysis of multi-sensor collaborative change trend,the health management of the whole life cycle of the wing is carried out,and the trend of faults is detected early,and the maintenance cost of the wing is greatly reduced.